Vasomotor Responses of Rat Coronary Arteries to Isoflurane and Halothane Depend on Preexposure Tone and Vessel Size

Methods: Thirty-six large coronary arteries (262 plus/minus 23 micro meter) and 42 small coronary arteries (99 plus/minus 15 micro meter) from 31 Wistar rats were isolated. Each vessel was placed in a microvessel chamber and was (1) submaximally preconstricted with the thromboxane analog U46619;(2) submaximally predilated with sodium nitroprusside; or (3) neither preconstricted nor predilated. The vessel was then subjected to increasing concentrations of either isoflurane or halothane, 0–3%. Changes in inner diameter were monitored and recorded with optical density video detection system.

Results: Isoflurane constricted predilated or untreated small coronary arteries, but had no effect on preconstricted small arteries. Isoflurane dilated large coronary arteries, with the preconstricted vessels dilating the most. In contrast, halothane dilated both the small and large coronary arteries to a similar extent. Preconstricted vessels dilated more to halothane than vessels with no added tone.

Conclusions: Whereas isoflurane has a heterogeneous vasomotor effect in rat coronary arteries, constricting the small vessels and dilating the large ones, halothane dilates both the small and large arteries. The vasoconstriction effect was most evident in vessels with no added tone, whereas the vasodilatory effect was most significant in preconstricted vessels. (Key words: Anesthetics, volatile: halothane; isoflurane. Coronary circulation: microcirculation. Vasomotor effect: heterogeneous.)

REIZ et al. first suggested that isoflurane induces coronary vasodilation and, under appropriate circumstances, causes redistribution of myocardial blood flow contributing to development of regional myocardial ischemia. As summarized in a recent editorial, subsequent studies have confirmed that isoflurane is a coronary vasodilator. Recently, however, we reported that in rabbits, isoflurane has a heterogeneous vasomotor effect on coronary arteries in vitro, constricting small subepicardial resistance arteries and dilating larger epicardial conductance arteries. The novelty of isoflurane-induced vasoconstriction of small coronary arteries raised the question of whether the finding was dependent on the unique experimental setup and/or species used and what vasomotor effect other volatile anesthetics might exert. .

In the current investigation, we extended our study to a second species, namely rats. In addition, we examined whether vessel tone before exposure to isoflurane may have an effect on subsequent vasomotion. This addresses one of the differences between our methods and previous in vitro studies in which the vessels or vessel strips were preconstricted with an agonist before exposure to a volatile anesthetic. Finally, the vasomotor effect of a second volatile anesthetic, halothane, is examined and compared to that of isoflurane.

In accordance with institutional Animal Care Committee standards, Wistar rats of either sex, weighing 100–150 gram, were anesthetized by 40 mg/kg ketamine injection and 5 mg/kg intraperitoneal xylazine. The heart was excised and coronary arteries were prepared as described previously. Two size groups of vessels were obtained--the epicardial left anterior descending arteries (N = 36, size 262 plus/minus 23 micro meter, range 220–336) and the subepicardial third or fourth generation branches in the left anterior descending distribution (N = 42, size 99 plus/minus 15, range 63–118). Each vessel was placed in a microvessel chamber, cannulated with dual glass micropipettes (50–100-micro meter diameter), and secured with 10–0 nylon monofilament sutures. The vessel was continuously bathed with modified Krebs buffer, gassed with 95% Oxygen²/5% Oxygen²mixture, and maintained at 37 degrees Celsius and a pH of 7.4. Oxygen tension (P^O2) in the vessel chamber exceeded 400 mmHg. The total volume of Krebs buffer circulating in the vessel chamber, buffer reservoir, and the connecting tubing was 100 ml. As the vessel was studied in a no-flow state, the pressure in the micropipettes was maintained at 40 mmHg to provide distention. The vessel was seen and its inner lumen diameter was measured and recorded, as described previously. In a preliminary study, the vessel diameter equilibrated within 5–10 min and remained stable with neither spontaneously vasoconstrictive nor vasodilatory tendency for at least 2.5 h. Vasomotor responses to KCl, the thromboxane analog U46619, acetylcholine, or isoproterenol were unchanged after 30 min in the vessel chamber versus after 90 min.

After equilibration of each vessel for at least 30 min in the vessel chamber, a baseline inner diameter was measured. The vessel was then randomized to be either (1) submaximally preconstricted by 15–20% of the baseline diameter by the thromboxane analog U46619 (0.22 plus/minus 0.18 micro Meter dose) in the vessel chamber;(2) sub-maximally dilated by 5–10% of the baseline diameter by sodium nitroprusside (0.97 plus/minus 0.17 micro Meter dose); or (3) neither preconstricted nor predilated. The vessel was subjected to increasing concentrations of isoflurane, 0.5%, 1%, 2%, and 3% for 10 min each, or halothane, 0.5%, 1%, 2%, and 3% for 15-min each, by adding the anesthetic to the 95% Oxygen²/5% CO²mixture bubbling the Krebs buffer solution, using an in-line Vernitrol bubble-through vaporizer (Ohio Medical Products, Madison, WI). Because anesthetic potency does not necessarily represent vasomotor potency, we elected to use equimolar amounts of the volatile anesthetics, rather than equi-minimum alveolar concentration amounts. The concentration ranges used represent approximately 0.4–2.6 minimum alveolar concentration for isoflurane and 0.6–3.9 minimum alveolar concentration for halothane, both of which are clinically meaningful. In a preliminary experiment, it was determined by gas chromatography that it took less than 10 min for isoflurane and 15 min for halothane to reach steady-state concentrations after introduction in the vessel chamber. The anesthetic content in the gas mixture was continuously monitored using a Rascal II Gas Analyzer (Ohmeda, Salt Lake City, UT), that had been calibrated with industrial standards. Selected samples were also taken from the vessel chamber to measure the concentration of isoflurane or halothane by gas chromatography. The millimolar concentrations and partial pressures of isoflurane (approximately 0.15–1.1 mM and 3.0–21.6 mmHg) and halothane (approximately 0.15–1.1 mM and 2.9–21.4 mmHg) remained consistently proportional to their concentrations in the gas mixture bubbled into the buffer solution. At each concentration of anesthetic, the inner diameter of the vessel was measured. At the end of each experiment, the anesthetic was discontinued. The vessel chamber was flushed with fresh Krebs buffer and the vessel was reequilibrated at 37 degrees Celsius. Potassium chloride was then added to achieve a concentration of 100 mM in the buffer and vessel chamber. Only those vessels that constricted by at least 15% to KCl were considered still viable and included for data analysis. This represented exclusion of any vessel that constricted less than the average by more than approximately 1 SD. Seventy-eight vessels from 31 rats met this criterion and are the subject of this study. No animal contributed more than one vessel to any experimental group; therefore, N for each experimental group is equal to the number of animals as well as the number of vessels.

Concentration-response curves of rat coronary arteries to increasing concentrations of isoflurane or halothane with different preexisting vessel tones were compared by multiple analysis of variance with one repeated-measures factor. Whether any group of vessels responded in a concentration-dependent manner to increasing concentrations of isoflurane or halothane was analyzed by a one-way analysis of variance (Scheffe's linear contrast). Significance was taken at P < 0.05. All statistics were calculated using True Epistat software (Epistat Services, Richardson, TX).

The vasomotor effect of isoflurane on the small sub-epicardial arteries depended on their preexposure tone (P < 0.005). Those vessels that were predilated and then exposed to isoflurane (N = 7, size 96 plus/minus 15 micro meter (mean plus/minus SD), range 75–115) and those that were neither predilated nor preconstricted (N = 6, size 93 plus/minus 20 micro meter, range 63–113) constricted in a concentration-dependent manner in response to isoflurane (P < 0.001 for each group). However, those that were preconstricted (N = 7, size 94 plus/minus 18 micro meter, range 71–118) did not dilate or constrict in response to isoflurane (P > 0.9)(Figure 1and Table 1).

The vasomotor effect of isoflurane on the large epicardial arteries also depended on their preexposure tone (P < 0.01). All three groups (predilated: N = 6, size 271 plus/minus 19 micro meter, range 253–304; preconstricted: N = 6, size 267 plus/minus 12 micro meter, range 244–280; neither: N = 6, size 285 plus/minus 30 micro meter, range 244–336) showed mild dilation in response to isoflurane (P < 0.05, 0.01, 0.01, respectively), with the preconstricted vessels dilating the most (Figure 2and Table 2). The effect of isoflurane on predilated large vessels was small in magnitude and probably unimportant physiologically.

The vasomotor effect of halothane on the rat coronary arteries also depended on the preexposure tone (P < 0.001 for small and for large arteries). Small subepicardial arteries that were preconstricted (N = 8, size 104 plus/minus 10 micro meter, range 88–117) and those with no added tone (N = 7, size 104 plus/minus 8 micro meter, range 85–113) showed concentration-dependent dilation to halothane (P < 0.001 for each), whereas those that were predilated (N = 7, size 99 plus/minus 11 micro meter, range 80–114) did not dilate in response to halothane (P = 0.26)(F1-23and Table 3). Similarly, large epicardial arteries that were pre-constricted (N = 6, size 251 plus/minus 17 micro meter, range 231–281) and those with no added tone (N = 6, size 250 plus/minus 12 micro meter, range 233–266) dilated in response to halothane (P < 0.001 and 0.05, respectively), whereas those that were predilated (N = 6, size 247 plus/minus 15 micro meter, range 220–269) did not dilate further (P > 0.9;Figure 3and Table 4). Preconstricted small subepicardial arteries did not dilate in response to halothane any more or less than preconstricted large epicardial arteries (P = 0.22).

The most important findings of this study were:(1) In vitro isoflurane is a selective vasoconstrictor of small subepicardial resistance coronary arteries in a second species, namely rats. (2) This effect is evident in predilated or untreated vessels, but is masked in submaximally preconstricted vessels, which may be the normal vasomotor state. (3) Isoflurane is a moderate vasodilator in large epicardial coronary arteries. This effect is greatest in preconstricted vessels. (4) Halothane appears to be a moderate, equally effective dilator of both small and large rat coronary arteries in vitro, being more evident in vessels with greater preexposure tone. There was no selectivity of the effect of halothane with respect to vessel size.

Our findings answer some of the questions raised by the novelty of the finding of isoflurane-induced vasoconstriction of rabbit small subepicardial vessels. First, the heterogeneous vasomotor effect of isoflurane--especially isoflurane-induced vasoconstriction of subepicardial resistance arteries--is not limited to rabbits, but also is found in rats, suggesting that it may be a more generalized phenomenon. Second, this effect can be masked by preconstriction of the small arteries. This may partially explain why previous in vitro studies that employed preconstriction of the vessels have not found isoflurane to have any constrictive effects. However, unlike previous in vitro studies in which isoflurane dilated preconstricted vessels or vessel strips, we found that isoflurane did not dilate or constrict preconstricted small arteries significantly.

Previously, advantages and limitations of our experimental preparation have been discussed extensively. Several issues merit further discussion, however. First, we measured the inner diameter of the vessels rather than isometric tension. Vascular resistance is a function of the vessel diameter and only indirectly of the tension of the vascular smooth muscle. Therefore, we submit it is more physiologically relevant to measure the vessel diameter than the vessel tone in studies of vascular resistance. In addition, magnitude of change in diameter may not have a 1:1 correlation with percent change in isometric tension. The latter, by definition, is tension with no change in dimension. Studies that measure changes in isometric tension may thus not be comparable to studies that measure changes in vessel dimension.

Second, the preparation employed in this study suspends the vessel in a no-flow state. The advantage of this is that we avoid flow-mediated release of endothelium-dependent nitric oxide, which may complicate experimental findings. This property is shared by most in vitro preparations of vessels and vessel rings. Data obtained in our preparation are equally valid as in other studies of in vitro vessel preparations.

These in vitro studies exclude the effect of the autonomic nervous system and blood-borne vasomotor mediators as well as autoregulatory and metabolic influences. This allows observation of the direct vasomotor effect of anesthetics. In contrast, it is a formidable challenge to validly document the direct effect of anesthetics on small resistance arteries in vivo. In vivo techniques and preparations incorporate a large variety of confounding variables. These include basal anesthesia, the use of calculated vascular resistance or coronary flow as a surrogate for small resistance artery vasomotion, measurement of epicardial collateral coronary vessels rather than true coronary arteries, unmeasured changes in compressive resistance associated with myocardial contraction, vasomotor effect of changes in metabolism, and autoregulation associated with hemodynamic changes. These factors and others make in vivo measurement of the direct effect of anesthetics on vasomotion problematic.

Conzen et al. measured the effect of inhalational anesthetics on the diameter of canine epicardial arteries of the in situ beating heart. These arteries measured 20–450 micro meter at baseline. Their technique represents a variation of stroboscopy developed by Nellis et al. and Marcus and coworkers. Nellis et al, applying the technique to rabbit hearts, noted that small vessels that were truly epicardial were veins, as verified by measurement of intravascular pressure. Small arteries tended to be subepicardial or progress along the epicardium for only short distances before penetrating into the myocardium; therefore, they could not be seen. A canine epicardium, studied by Conzen et al., represents an exceptional case in that dogs have extensive epicardial collateral arteries of varying sizes. Conzen et al. made no effort to distinguish between true resistance arteries and arterioles on the one hand and collateral arteries of similar caliber on the other. Collateral arteries have less developed smooth muscle and poorer contractility than normal arteries. Furthermore, in other species such as pigs and humans, epicardial collateral arteries are poorly developed. Results obtained with collateral arteries may not be applicable to true resistance arteries.

Whereas none of the in vitro or in vivo methods are ideal or without limitations, the in vitro video detection measurement of isolated vessels suspended in a no-flow state represents a valid method to study the direct vasomotor effect of anesthetics.

Merin, in a closed-chest dog preparation, observed proportional decreases in myocardial blood flow (MBF) and oxygen uptake with increased isoflurane concentration. Cheng et al. described unchanged coronary resistance in an isoflurane-anesthetized swine model. These two studies contrast with many others that describe decreased coronary oxygen extraction, equated with isoflurane-induced vasodilation, whether MBF is increased, decreased, or unchanged. Recent studies by Larach et al. and Crystal et al. have documented isoflurane-induced concentration-dependent increases in MBF in the absence of indirect effects from autoregulation, metabolic changes, and compressive resistance. However, an increase in MBF may represent dilation of coronary resistance vessels and/or increase in coronary nonnutritive shunting. Merely measuring changes in MBF does not indicate the vascular site of isoflurane-mediated effect.

The question of whether isoflurane can increase coronary nonnutritive shunting has not been answered adequately. Gelman et al. used two different sizes (15 micro meter and 9 micro meter) of microspheres to measure the total and nutritive flows and calculated nonnutritive shunt as the difference between the two. They noted a threefold increase in intramyocardial shunting in dogs under 2 minimum alveolar concentration isoflurane anesthesia, although the increase did not achieve statistical significance. Crystal et al., using 15-micro meter microspheres only, reported that under both control and isoflurane anesthesia, arteriovenous shunting remained at less than 2%. According to Gelman et al., however, one needs to measure both total flow (nonentrapment of 15-micro meter microspheres) and nutritive flow (nonentrapment of 9-micro meter microspheres), using two sizes of microspheres, to calculate nonnutritive shunt. Finally, a decrease in oxygen extraction and an increase in coronary sinus oxygen content under isoflurane anesthesia would be consistent with an increase in nonnutritive shunting.

Previous in vitro studies of coronary vasomotor effects of volatile anesthetics examined preconstricted segments of conductance arteries. They observed that isoflurane decreased tension of such preparations. This is consistent with our observation that in large epicardial arteries, isoflurane-associated vasodilation is especially evident in preconstricted vessels. Whereas in vitro studies of conductance vessels allow isolation of the direct vasomotor effect of the anesthetic, conductance vessels contribute only a small percentage of total coronary resistance under normal circumstances. The small subepicardial arteries we have studied are sufficiently small to be considered true resistance vessels. These resistance vessels demonstrate isoflurane-induced vasoconstriction in the absence of added tone.

Differences in vasomotor responses to isoflurane among different portions of the coronary circulation have been noted before. Nakamura et al. compared the proximal and smaller distal portions of the left anterior descending and left circumflex arteries in dogs. They noted that 3.5% isoflurane dilated the distal portions more than the proximal portions, whereas lower concentrations had no differential effect. Even at the highest concentration, the dilatory effect was moderate at best. Even the small arteries of Nakamura et al. are not true resistance arteries, but conductance arteries. Our conductance arteries demonstrated mild to moderate dilation to isoflurane as well. The mechanistic basis for heterogeneous vasomotor responses among different portions of the coronary circulation has not been investigated and requires further study.

Our finding of heterogeneous vasomotor effect of isoflurane, with vasoconstriction of small resistance arteries and dependence on preexposure tone, predicts the following effects on coronary resistance and flow in the normal coronary circulation. In this situation, the coronary vessels possess a significant amount of vasomotor tone and autoregulatory reserve. Thus, we speculate that preconstricted vessels may best approximate the vasomotor state of normal coronary vessels. The direct effect of isoflurane would be mild dilation of the large conductance arteries and no significant change in the vasomotor state of the small resistance arteries. Because the small arteries account for the great majority of the total resistance even without isoflurane, the net change in vascular resistance from isoflurane is likely to be small. Rather, competing influences of autoregulation and metabolism-flow coupling may have greater effects on the total resistance and flow. Indeed, whereas some studies report a large increase in myocardial flow with isoflurane, most report little change in flow to the normal myocardium. .

The effects of isoflurane on the regional coronary circulation supplied by a hemodynamically significant or critical stenosis and on the regional coronary circulation supplied by collateral blood vessels must be complex. Further studies are needed to resolve controversies raised by previous studies and to define the direct and net effects on myocardial perfusion.

Unlike isoflurane, we have found that halothane dilates both the large epicardial and the small subepicardial arteries, with no selectivity. The vasodilatory effect of halothane was greater in preconstricted arteries than arteries with lesser preexisting tone. Two previous in vitro studies, performed with preconstricted porcine large coronary artery segments, demonstrated vasodilatory effect of halothane. Our finding in rats is consistent with theirs. In addition, Nakamura et al. demonstrated in dogs that halothane dilated proximal coronary arteries (outer diameters of 2.5–3.2 mm) more than distal arteries (0.6–0.9 mm). This preferential effect was not observed between our size groups of arteries. Differences in the findings may be due to differences in the sizes of vessels studied, other aspects of experimental methods or species studied.

In vivo studies of the vasomotor effect of halothane in coronary arteries indicate that its indirect effect via depression of myocardial metabolism and metabolism-flow coupling may be at least as important as its direct vasodilatory effect. Doyle et al. observed in open-chest dogs that halothane caused a dose-dependent decrease in MBF, along with a similar reduction in contractility. Moore et al. demonstrated that whereas 0.4% halothane caused no change in myocardial oxygen consumption and MBF, 1.2% halothane decreased both. In contrast, when anesthesia was induced in dogs with halothane and myocardial work was allowed to increase with autonomic activation, MBF increased. With autonomic blockade, halothane induction did not change MBF. Finally, in tetrodotoxin-arrested rat hearts, in which myocardial oxygen consumption is constant, Larach et al. demonstrated that halothane increased MBF in a dose-dependent manner.

In summary, we have shown that isoflurane has a heterogeneous vasomotor effect in an in vitro coronary arterial preparation of a second species, namely rats. It causes vasoconstriction of small resistance arteries, but dilates large conductance arteries. The vasoconstrictive effect is most evident in small arteries with no added tone, whereas the vasodilatory effect is the strongest in large arteries with preconstriction. In contrast, halothane dilates both the large and the small arteries to a similar degree. This effect is most evident in preconstricted arteries. The net effect of the anesthetics on coronary flow will be a function of not only their direct effects demonstrated in this study, but also of the indirect effects on myocardial metabolism and metabolism-flow coupling, myocardial contraction and compressive resistance, and systemic hemodynamics and autoregulation.